Abstract Thiopurines (e.g., mercaptopurine [MP]) are highly effective against hematologic malignancies (leukemia and lymphoma) and autoimmune diseases (inflammatory bowel diseases [IBD]). In acute lymphoblastic leukemia (ALL) in particular, prolonged daily exposure to MP is indispensable for the cure of this cancer. However, thiopurines have narrow therapeutic indexes with dose-limiting hematopoietic toxicity that causes extensive morbidity, disruption of treatment, and possible mortality. Therefore, there is enormous clinical benefit from preemptively identifying patients at risk of thiopurine toxicity and individualizing therapy to mitigate it. Thiopurine toxicity is also highly influenced by inherited genetic variations, particularly polymorphisms in the TPMT gene as discovered by our group and others. Recently, we reported a novel variant in the NUDT15 gene that lead to its loss of nucleotide diphosphatase activity, excessive levels of active thiopurine metabolite, dramatic increase in MP-induced apoptosis and severe toxicity in patients (J Clin Oncol 2015 and unpublished preliminary results). Because the low activity NUDT15 allele alters the metabolism of thiopurines, we hypothesize that we can rationally reduce thiopurine dose in patients who inherit the NUDT15 variants and tailor their exposure to thiopurine active metabolite TGTP to the level comparable to wildtype patients receiving conventional doses, similar to the principle of TPMT-guided thiopurine dose reduction. To test this hypothesis, we propose three specific aims to 1) comprehensively identify MP toxicity-related NUDT15 and TPMT ALL (N=1,028), 2) characterize the effects of NUDT15 and characterize how NUDT15 and variants influence thiopurine disposition in children with ALL (N=1,550), TPMT variants variants in children with on its function, and finally 3) TPMT from which to develop a NUDT15-/TPMT-based pharmacogenetic algorithm for MP dose adjustments. Successful completion of these studies is likely to establish a novel precision medicine paradigm for thiopurine therapy to proactively individualize dose before toxicity occurs. We are confident that this highly translational project will likely have immediate impact on the treatment of ALL, and our findings can be readily extrapolated to thiopurine therapy for non-malignant conditions (e.g., IBD) thus impact a large number of patients.